Re: Re[2]: [gnugo-devel] life & death (3.3.5)

[Gunnar Farneback]

Paul wrote:
>    - eye-space analyzing in life.c and optics.c doesn't consider the

You can safely ignore the code in life.c. It is not in use and is
likely to be removed sometime soon.

>      fact that strings at A4 and at C7 (as well as at A8 and at C7)
>      are not connected. linear five vertices eye shape is considered
>      to give 2 eyes unconditionally (per eyes.db).

Correct, the eye analysis doesn't have the concept of cutting points.

>    - owl reading doesn't go deep enough to find that A5 is a critical
>      point.

It's not really a question of depth but rather that it doesn't
understand that the move is worth investigating.

> i think we could implement the following concept:
>    - list all the strings the dragon consists of (A8, C7, A4).

This step is easy. :-)

>    - check if they are connectable with moves outside eye shape (A8 and
>      C7 can be connected by C8, A4 and C7 cannot be connected).

This is not trivial, in the sense that we don't have a machinery in
place for this analysis.

>    - separate the dragon into parts which must be connected (in case
>      of attack) with moves inside eye shape (2 parts A8+C7 and A4).

This is not clear. If black attacks at C8, doesn't white have to
connect A8 and C7 inside the eye space as well?

>    - enumerate common liberties/connections of different parts (here
>      goes the vague part) - parts can be "connected" by A5 or B5.

Indeed, this is quite vague.

>    - note that black A5 forces us to play B5 sooner or later (black
>      can just make atari).

This is kind of a key point, but how do we carry out the analysis
leading to the conclusion?

> [...]

> well, this is an intricate explanation, but i think i managed to
> formulate the main idea. maybe it would allow to impove life & death
> code somewhat.

It's a good start, even though it probably needs a lot of refinement.
In any case this is an important problem to solve.

/Gunnar